MICHELLE DELIO▊
- BYLINE
- Wired News · MIT Technology Review · InfoWorld (also filed as "Michelle Finley")
- BEAT
- Hacking, security, tech culture — where the sources are anonymous and the readers cannot check them
- METHOD
- The email interview; the source used once and not kept — her own stated practice [3]
- FINDING
- Independent review: 3 of 10 examined Technology Review stories had sources whose existence could be verified, April 2005 FACT
- STATUS
- SCRUBBED FROM THE RECORD — Technology Review removed all ten stories it examined FACT
- DISPOSITION
- Wired News amended 24 stories; InfoWorld four. No criminal charge — a journalism-ethics matter, not a crime FACT
The specimen, observed roughly 1998–2005 across the tech-press canopy, filed some seven hundred stories on a beat where the sources could not be checked. When the field finally checked, most of the disputed ones were not there.
What you are about to see is not a hack, a fraud conviction, or a sanction. It is quieter than that, and in its way more instructive: a working reporter who chose a beat where the sources hide by nature, and filled that beat with sources no one else could find.
The publications went back and checked. That is why the archive exists. Every event below carries its receipt. Both sides get the microphone — and her defense, which is a real one, gets it at full volume.
the drama timeline
ACT I — THE VOLUME (1998–2004)
The output comes first. It always does. Watch the byline multiply across three mastheads until no single editor can hold the whole of it in view — which is another way of saying no single editor can audit it.
-
1998–2004
Seven hundred bylines
Delio files prolifically for Wired News, and also for MIT Technology Review and InfoWorld — a body of work later put at more than 700 Wired News stories. The beat is hacking and security, where the best sources are anonymous by profession and the reader has no way to run them down.
A beat perfectly shaped for undetectable invention: the insider who will only speak unnamed is the one witness a fact-checker can never quite reach.
-
2001
"Wag the Delio"
Attrition.org, tracking the security press, coins the phrase "Wag the Delio" around a Black Hat 2001 presentation, characterizing her reporting on a supposed U.S.–China hacker "cyberwar" — built, in attrition's telling, on IRC chatter dressed as intelligence. Attrition and Gelf Magazine also note a prior career unusual for the beat: professional palm-and-tarot reading, and editing Outlaw Biker magazine — color that appears in those write-ups, not in the later Technology Review or Wikipedia record.
ACT II — THE UNRAVELING (2005)
One challenged source is a coincidence. The second is a question. By the time a journalism school is counting, it has become a method.
-
FEB–MAR 2005
Hewlett-Packard raises the alarm
Hewlett-Packard flags a Delio piece for Technology Review, "Carly's Way," reporting an anonymous insider's account of how then-CEO Carly Fiorina had been treated. HP cannot locate the story's sole named source, "Keith Abrams." Technology Review retracts two pieces in March and opens a formal review.
-
APR 20, 2005
The Berkeley count
Technology Review publishes the findings of an independent review led by UC Berkeley journalism faculty member Susan Rasky with five graduate students. Of ten stories examined, three had verifiable sources; the rest ranged from partially unverifiable to, in one case, a quote military sources denied giving. Technology Review removes all ten — including the three that checked out.
Note the editor's arithmetic. Once enough of the sources are phantoms, the burden inverts: the real ones come down too, because the byline can no longer vouch for any of them.
-
MAY 9–12, 2005
Wired counts too, and so does InfoWorld
Wired News runs its own review — commissioning NYU journalism professor Adam Penenberg, who had exposed Stephen Glass at The New Republic in 1998. Wired amends 24 stories whose sources it could not identify; InfoWorld amends four. The disputed sources are, consistently, providers of "color" — atmosphere quotes rather than load-bearing facts.
ACT III — THE DEFENSE (2005)
Cornered organisms explain themselves. This one did not litigate and did not vanish quietly; it made an argument — and the argument is worth hearing, because it is the argument every unverifiable byline will make.
-
MAY 16, 2005
Delio answers, in her own words
Technology Review publishes her response. Her account: most of her interviews were done by email, and she did not retain messages from sources she did not expect to use again — "I don't understand why my credibility and career are now hanging solely on finding minor sources that contributed color quotes to stories I filed months and years ago." Going forward, she said, she would keep a file on everyone quoted.
The defense rests on a true premise — nobody archives every email — carrying a heavier load than a true premise can bear. Missing email explains one lost source. It explains a pattern less easily.
-
EPILOGUE
The byline goes quiet
After the 2005 reviews, the prolific tech-press byline effectively ends. Technology Review's removed stories stay removed; Wired's amendments stay amended. The attrition dossier remains online. The archive keeps the tape running.
both sides, on the record
The independent finding: Technology Review's commissioned review, led by UC Berkeley's Susan Rasky, could verify the sources of only three of ten examined stories; the publication removed all ten [1] [2].
The admission: in one case, per Technology Review, Delio admitted she had manufactured a quotation — the one fact that turns "sloppy" into something else [2].
The pattern: Wired News amended 24 stories after its own review; InfoWorld amended four. The problem spanned three mastheads and several years, not a single deadline [4].
Delio maintained her sources were real. Her stated position: interviews were done largely by email, and she did not keep messages she did not expect to reuse, which left her unable to produce them months or years later — not proof they never existed [3].
What was and was not in dispute: the challenged sources were, by the reviewers' own description, providers of "color" — and the main substance of the stories was not shown to be false. Three of the ten Technology Review stories verified cleanly and were pulled anyway [5] [1].
Scope of the record: no court, licensing body, or agency ever adjudicated any of this. It is a journalism-ethics dispute resolved by editors, not a crime; the "fabrication" finding is Technology Review's editorial judgment, and the amendment counts are the publications' own [2].
YOU DECIDE
Scoped to the finding, never the person. What is settled: an independent review verified the sources of only 3 of 10 examined stories, and Technology Review reported that Delio admitted manufacturing one quotation. Those two, she has not rebutted. The wider charge — that all seven hundred bylines were invention — nobody proved, and this page does not assert it.
Weigh the costly signal: three publications reopened years of her filed work and put their own names behind the corrections. Her answer was that she does not keep her email.
The archive does not judge. The archive merely keeps the tape running.
evidence locker
PRIMARY RECORD
PRESS & DISCOURSE
WATCHDOG & REFERENCE
The standard. Everything above is sourced to the publications' own investigation pages, named press reporting, the subject's own published response, and an archived watchdog file. Facts are stated as facts; the "fabrication" finding is stated as Technology Review's editorial judgment; attrition.org's reads are stated as attrition's. The defense is presented at full strength, including what the record does not show — no court ruled on any of this, and the disputed sources were "color," not the substance. No motive is asserted, no private character diagnosed. The burden of proof is on us, not the subject. If it couldn't survive a defamation challenge, it wouldn't be on this page.